The purpose of this research program is to delineate more fully how auditory and visual cues are combines in bisensory speech perception by hearing-impaired subjects, and to identify strategies to improve auditory-visual (AV) recognition performance for individuals whose AV scores are less than optimal. In support of these objectives, the proposed research will (1) determine the most useful auditory cues for optimizing auditory-visual performance, (2) evaluate the ability of subjects to integrate A and V es at both the segmental and connected speech levels, and (3) evaluate the ability of subjects to use contextual constraints to aid in the recognition of words in sentences. Important aspects of this proposal include the assessment of auditory-visual integration ability at both the segmental and connected speech levels, and the modeling overall auditory-visual performance as a combination of three sub-components: (1) auditory recognition ability of relevant acoustic cues, (2) auditory-visual integration ability, and (3) the ability to use context. In addition, performance variability among hearing-impaired subjects on each of these three factors will be assessed, and the value each factor as a predictor of overall AV benefit will be determined.